=== chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk [06:51] good morning [07:03] good morning === chihchun_afk is now known as chihchun === rsalveti_ is now known as rsalveti === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === wesleyma` is now known as wesleymason [08:09] shrug [08:09] u-d-f on personal fails with a "failed to find user uid/gid" :-/ [08:14] sergiusens, ^ do you have any idea about that one? [08:14] i guess half the world is off today [08:14] why? [08:14] * ogra_ cant remember if sergio was among them [08:15] is today some religious holidays most country have or something? [08:15] independence day in the US ... and there was something in south america i cant remember [08:15] hum [08:15] isn't independance day the 4th of july? [08:15] e.g tomorrow [08:15] yep [08:15] so to make it worth it they get today off ... funny US law [08:16] that's cheating! [08:16] (because of weekend ... but having the right for a free day) [08:16] is mvo celebrating US independance as well? ;-) [08:16] yeah, totally ! [08:16] haha, for sure [08:16] slacker! [08:16] yeah [08:17] seb128, are you seeing that error when building the image or when booting it ? [08:19] seb128: yet you don't think it's cheating when we get friday and Monday off for xmas that fell on the saturday or sunday right ;) [08:20] davmor2, "we"? [08:20] so [08:20] 1- France doesn't have that rule (nor does most european countries I think) [08:20] 2- yes, I find it cheating when the u.k do that [08:21] ogra_, u-d-f [08:21] bah [08:21] I guess it's some of the image processing [08:22] seb128, well, thats like being in the EU but not having the euro, so what do you expect :) [08:22] dunno if that has to do with the cloud-init cleanout [08:22] that string isn't in goget-u-t though [08:23] * ogra_ waits for the 44min download to finish :P [08:23] Hello everybody [08:23] I have a BBB I don't know how to drive a pwm, using snappy. I could do this before, using angstrom distribution. [08:24] I know the device tree overlay is different with angstrom, but i don't know how can I configure the hardware in snappy. [08:24] Can anyone help me? [08:24] ogra_, http://paste.ubuntu.com/11814398/ [08:24] bah [08:25] Rlyeh, we dont really have a standard way for overlay dtb files yet ... [08:26] there was a long thread about it on the mailin list (snappy-devel i think) [08:26] *mailing [08:26] You mean I can't use all hardware resources? [08:27] we have no standard way, you need to hack in the overlay currently (i think it is described in the thread) [08:27] I asked there, but no one didn't answered me [08:28] It's just about BBB or it's same for RPi2 too? [08:28] bah [08:28] ogra_, https://bugs.launchpad.net/ubuntu/+source/goget-ubuntu-touch/+bug/1470727 [08:28] Ubuntu bug 1470727 in goget-ubuntu-touch (Ubuntu) "ubuntu-device-flash touch fails with: failed to find user uid/gid" [Undecided,Confirmed] [08:29] so not only personal [08:29] likely one of the ubuntu-snappy changes from this week I guess [08:29] Rpi is a little worse since it uses the binary blob to initially set the dtb's ... which is the reason we dont have a standard way yet ... that standard way needs to cover both implementations [08:29] since on my box I didn't upgrade anything [08:29] so it's not udf/the host system versions [08:30] might be ... lets see if i can verify here [08:30] thanks [08:30] (in about 40min ) [08:31] OK, thank you anyway ogra [08:31] Rlyeh, i'm partially working on the RPi implementation, once there is a proper way i'll send a mail to the list (and put up documentation) [08:32] Thank you, wish you be success in that [08:34] ogra_, you can try on core if you want, less to download and it gives the same error [08:34] grrr [08:35] wtf snappy team, wrecking stuff on a virtual-friday [08:36] that string is part of /usr/bin/snappy [08:37] I guess it's due to rsalveti changes yesterday [08:37] http://launchpadlibrarian.net/210553123/ubuntu-snappy_1.3ubuntu1_1.4ubuntu1.diff.gz [08:37] lot of uid things in there [08:38] or... [08:38] I might need to update ubuntu-snappy on my system [08:39] oh, snappy vs clickpkg again [08:44] ogra_, it works after installing the new ubuntu-snappy on my laptop [08:44] so yeah, mismatch between the tools and the image [08:44] yeah [08:45] because it only finds the snappypkg user [08:45] (or rather doesnt find it) [08:48] seb128: the uid/gid thing might be because we renamed a user [08:48] seb128: might :) [08:48] Chipaca, seems to be yeah [08:56] ok, works [08:56] today personal has a working ubuntu user without cloud-init, things are improving ;-) [08:57] well, still rebootloop by default here [08:58] right, the grub issues are not resolved [08:58] you need to pick the "ubuntu" entry [08:58] but it's not a surprise since those changes didn't land [09:00] hmpf, why do i land in emergency mode now [09:10] Good morning all; happy Friday, and happy International Plastic Bag Free Day! 😃 [09:14] seb128, so any idea why the ubuntu user doesnt show up in lightdm ? [09:18] ogra_, no, it seems random/transient, I got it showing up after systemctl restart lightdm and it was there on next boot as well [09:19] ogra_, I think I'm going to change the image to autologin [09:19] yeah [09:19] we ship a snippet for that in livecd-rootfs for the phone [09:19] live-build/ubuntu-touch/includes.chroot/etc/lightdm/lightdm.conf.d/90-phablet.conf [09:24] ogra_, any reason that's not in /usr/share/lightdm/lightdm.conf.d? [09:25] I guess location changed and that one didn't get migrated [09:25] yeah, there was one, ask mterry [09:26] * ogra_ doesnt get it ... i always end up in emergency mode now [09:26] what mode is that? [09:26] systemd drops you into a rootshell [09:27] rsalveti (and everyone else): I filed https://bugs.launchpad.net/snappy/+bug/1471160 to track the work we talked about - most of it is on developer.u.c obviously, but there's the thing about converting the .rst file in the docs to .md [09:27] Ubuntu bug 1471160 in Snappy "Import Snappy docs automatically into developer site" [Undecided,New] [09:27] davidcalle, ^ [09:30] hmm, so just dropping the autologin snippet from touch in place doesnt help [09:31] (indeed with s/phablet/ubuntu/) [09:37] dholbach, thanks, I've just found out how we can support markdown tables in an importer (md doesn't have tables by default, this was the reason for the rst) [09:37] oooh nice! [09:43] dholbach, I'm starting the rst->md conversion, I've tried a few automated things and they seem to fail on it [09:43] maybe it doesn't necessarily need to be a table? :) [09:46] dholbach, hmm, it uses them quite a lot : https://developer.ubuntu.com/en/snappy/guides/filesystem-layout/ [09:46] I see, mh [10:00] davidcalle, shall we have a quick chat now or would you prefer later? [10:00] dholbach, now is fine :) [10:00] perfect [10:00] let me set up the hangout === chihchun is now known as chihchun_afk === zyga_ is now known as zyga === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk [12:21] ogra_, did you mean earlier than systemd was booting in state "maintenance"? [12:23] ogra seb128: you need the latest snappy installed on your host for that to go away (in wily, I think rsalveti copy-package'ed it to snappy-dev/tools-proposed) [12:23] Chipaca: ^ [12:23] sergiusens, yeah, I figured that out/resolved it [12:23] thanks [12:23] sergiusens: which "that"? [12:23] * Chipaca out of the loop? [12:23] sergiusens, yep, we found already [12:24] seb128: from my point of view the image was broken all week and I spent most of it chasing that [12:24] ;-) [12:24] sergiusens, same here... [12:24] seb128, well, not sure how it is actually called (it boots normal now) but it talked about systemctl reboot and systemctl default in the text it printed before giving me a shell [12:24] k, well for me system doesn't boot [12:25] (note that i dropped all console= options so i can actually see error messages) [12:25] (and panic= as well) [12:28] ogra_: what does console default to without console=, in different arches? [12:28] tty0 [12:28] on all arches afaik [12:29] i would only set console= if i actually want serial ... (and only set it once) [12:29] for all other variants i'd just drop it [12:33] davidcalle: dholbach can you switch to github flavored markdown? https://help.github.com/articles/github-flavored-markdown/ [12:37] sergiusens, I don't know - with which tool can we convert it to html? [12:39] dholbach: that python one I forget the name of... Chipaca helpz :-P [12:39] yeah, ask sergiusens he once told me :P [12:41] ah, pandoc ... [12:42] sergiusens, do you need specific features we can't have for now? [12:43] Chipaca: btw, the find label was doing something useful [12:43] davidcalle: the tables ;-) [12:44] sergiusens, ah, then in that regard, yes, you will be able to use anything github md do for tables [12:44] s/will be/are [12:44] Chipaca: now I can't find the alternate kernel path because / is (hd0,2) and there is no kernel in /vmlinuz only in (hd0,3) and (hd0,4) which correspont to the searched for labels [12:46] Chipaca: I added back the search and it all works again :-/ [12:50] was afk (lunch etc) [12:50] * Chipaca reads === chihchun_afk is now known as chihchun [12:51] Chipaca: summary --set with no var set $prefix to the found $label [12:51] sergiusens: wait what? [12:51] sergiusens: ooooooh! but that's not what the docs say :) [12:51] ok [12:51] Chipaca: it does :-) [12:51] docs say it sets root [12:52] and that $prefix is calculated before all user-provided stuff? or something like that [12:52] Chipaca: hmm, I read that prefix defaults to the bootloader partition, in any case /vmlinuz alone points to nowhere :-P [12:53] Chipaca: https://code.launchpad.net/~sergiusens/snappy-hub/searchForMe/+merge/263772 [12:53] sergiusens: changing subjects, what do you think of .snapignore? [12:53] Chipaca: sorry, not prefix but root [12:54] ah, yes, that yes [12:54] but we don't use root [12:54] Chipaca: that is inherited from click, which I know because I proposed adding a couple of files there and cjwatson said no, and had this huge list :-P [12:54] Chipaca: yes we do [12:54] anyway, [12:54] if you tell me it works like this and does not without, screw the docs, this stays :) [12:55] Chipaca: it works [12:55] example from the docs: set root=(hd0,1); linux /vmlinuz [12:55] sergiusens: where do we use $root? [12:56] Chipaca: http://www.gnu.org/software/grub/manual/grub.html#root [12:56] Chipaca: when you use '/' without specifing the partition [12:56] Chipaca: as in vmlinuz and /initrd.img === chihchun is now known as chihchun_afk [12:56] sigh. ok. [12:56] */vmlinuz [12:57] yep yep [12:57] i blame ogra :) [12:57] he wants to confuse us, so we'll switch to uboot on intel [12:57] it's fine, he should of catched this [12:57] heh [12:57] we can I guess :P [12:57] it's because he has shares in the uboot corporation [12:58] sergiusens: wrt .snapignore [12:58] ~chipaca/snappy/dynamicExclusion [12:58] lp:^ [12:58] sergiusens: would something like that be potable? [12:59] sergiusens: http://bazaar.launchpad.net/~chipaca/snappy/dynamicExclusion/revision/555 i guess [13:03] Chipaca: do you reread the file for every file there? it looks good in any case [13:04] sergiusens: wanted to have it not break if somehow two builds are called in the same run (not something that can happen easily) [13:04] sergiusens: it rereads the file for every build directory [13:04] Chipaca: idea, how about we read /proc/net/dev as part of the first boot task and set the appropriate network interface up correctly? [13:04] +1 [13:05] and indeed, uboot would fix everything ! [13:05] sergiusens: i'll write tests for this, and add that to first boot [13:05] or i could do it the other way around [13:05] \o/ [13:05] * Chipaca shelves this [13:05] well, wait [13:05] pitti had said something about something changing? [13:05] wow, really? :-P [13:06] did he ? [13:06] we don't like change! [13:06] i might have misunderstood [13:06] yeah, we're no obama ! [13:06] * sergiusens mentions that he is being sarcastic in case people don't notice [13:06] pitti: this is wrt network device names [13:06] sergiusens, its friday ... everything is sarcastic [13:06] * ogra_ melts [13:06] pitti: is the current situation "it"? or are we expecting more changes soon? === devil is now known as Guest54128 [13:07] pitti: because if it's "it", we know how to work with it :) [13:07] well, even if it changes ... /proc/net/dev should give us the right info [13:07] fair enough [13:08] ogra_: btw, do you know why we don't "see" the bbb usb net device? [13:09] missing module ? [13:09] tried usbnet, not it [13:09] no [13:09] its an smc1234something [13:09] o, that would be a great thing to work on as well [13:09] oh [13:09] you mean gadget support ! [13:09] now i get it [13:09] do we want to bring up the first ether device, or the first anything-but-wireless device? [13:09] * ogra_ needs some cooling heard ... brain only operates at 20% or so [13:10] *head [13:10] ogra_: massagran [13:10] no, not massagran [13:10] 'old on [13:10] Chipaca, we'd need to ship the cdc_composite module and configure it accordingly [13:10] ogra_: mazagran [13:10] ogra_: cold coffee + lemon ftw [13:11] hmm, sounds like an odd combo [13:11] 's good [13:11] but don't do the version with rum just yet [13:11] https://en.wikipedia.org/wiki/Mazagran_(coffee_beverage) [13:12] ogra_: how was it you asked to be dropped into initrd? [13:13] i just had a longish discussion with asac about that :) [13:13] (modules handling etc) [13:14] my idea is to have a modules.img in /boot ... (or multiple if one is to big) ... so you can loop mount that from initrd ... and later move mount it to the rootfs (no duplication, no modules inside the initrd) [13:14] i discussed that with rtg earlier already and will work on a prototype soon as proof of concept [13:14] WOW ... [13:15] susie just brought me a bowl of raspberries ... fresh from the bush [13:15] they are like cooked ... totally hot [13:15] Chipaca: I don't plan on changing the current situation myself, unless we have to because of bugs [13:16] Chipaca: we might get some exclusion rules for rmnet* devices for ofono; but that wouldn't affect eth or wlan === chihchun_afk is now known as chihchun [13:16] Chipaca: OOI, are you using networkd now, or just a more dynamic ifupdown snippet? [13:16] yeah, that would also only affect personal === chihchun is now known as chihchun_afk [13:17] where we wont have a hard configured eth0 or wlan0 anywa [13:17] y [13:17] pitti, no networkd yet [13:17] (afaik) [13:17] not on core, no [13:18] might be interesting to know how that works with ubuntu-fan [13:18] or if at all [13:18] sergiusens: how can i tell from first boot if i'm in personal or not? [13:18] or would networkd running mean no ifup [13:18] because if so we're good to go :) [13:18] Chipaca, i guess that can only be answered once personal exists :) [13:19] Chipaca, ogra_: right, I was just intersted (it's easier to configure that with networkd than with ifupdown, but I guess you already have a solution that writes interfaces on first boot or so? [13:19] pitti, not really [13:19] pitti: that's what i'd be doing, at least for core [13:19] livecd-rootfs currently creates a file [13:19] write the equivalent of the current static eth0 file on first boot [13:19] Chipaca: yeah, you shouldn't configure it in multiple places; pick one of ifupdown, NetworkManager, or networkd [13:19] Chipaca: ack [13:26] pitti: do usb network interfaces also start with /en/? [13:27] Chipaca: depends -- a tethered phone yes, but I figure an USB wifi stick would start with wl* [13:27] pitti: yeah, i meant something like a tethered phone, or a usb ethernet device, not a wifi nor wlan [13:27] things that would be expected to 'just work' for an iot device :) [13:28] Chipaca, usually usb0 etc [13:28] depends on the gadget driver you use [13:28] ogra_: before yes, but now? [13:28] heh [13:28] i guess we'll see [13:28] * ogra_ will tinker a bit with gadget stuff on the weekend [13:29] ogra_: http://cgit.freedesktop.org/systemd/systemd/tree/src/udev/udev-builtin-net_id.c#n20 [13:29] hmm [13:29] so ens3 is the ethernet device plugged into the hotplug slot number 3 [13:29] that has no usb at all [13:29] ogra_: nor should it [13:29] ogra_: i mean, unless you expect ethernet devices to mention they're pcie [13:30] well, you want to name them somehow [13:30] -- PCI geographical location [13:30] * [P]ps[f][u][..][c][i] [13:30] * -- USB port number chain [13:30] so usb would be enPsomethingpsomethingssomethingfsomethingusomethingcsomething [13:30] ah, i was reading the comment at the top [13:30] supereasy [13:31] which kind of indicates it will never even find a device on something like the BBB [13:31] * Chipaca is now very curious as to how this looks in actual practice :) [13:31] ogra_: why? [13:31] ogra_: it'll start with en [13:31] * Predictable network interface device names based on: [13:31] * - firmware/bios-provided index numbers for on-board devices [13:31] * - firmware-provided pci-express hotplug slot index number [13:31] * - physical/geographical location of the hardware [13:31] just then have a bunch of stuff [13:31] * - the interface's MAC address [13:31] the only thing that might apply is the MAC in that list [13:31] 's fine [13:31] as long as it starts with /en/, i'll bring it up :) [13:31] that code is so awfully intel centric ... sigh [13:32] bios ... pci ... [13:32] ogra_: it's written by people who were rather late to the arm bandwagon [13:32] not particularly surprising [13:32] it is written by people who tend to ignore everything but intel [13:33] which was/is one of my biggest concerns switching to systemd [13:33] init=/bin/sh ... really ... :P [13:33] ogra_: as i say, super interested in bringing up that bbb thingie, now :) [13:34] now ... [13:34] ogra_: init=/bin/snappy \o/ [13:34] or that, yeah ! [13:34] anyway, need to go for the school run. will bbl :) [13:34] check if the cdc_composite module exists [13:34] * Chipaca checks [13:35] or cdc_ether [13:35] (but that would limit you to only ethernet over USB) [13:36] hmm, the RPi kernel doesnt have it enabled :( [13:37] * Chipaca wonders what ttyO0 is [13:38] Chipaca: the O is from OMAP [13:38] https://www.kernel.org/doc/Documentation/usb/gadget_multi.txt [13:38] I guess [13:39] yeah [13:39] damjan: ta :) [13:39] which is weird since the BBB has no OMAP [13:39] but TI stuck to the O [13:39] ogra_: i don't have cdc_composite [13:39] Chipaca, the above text should help [13:39] ogra_: i do have cdc_ether [13:39] yeah, thats just therenet over usb ... not host mode though [13:40] *ethernet [13:40] yeah, no workie [13:40] you want actually g_ether [13:40] or g_multi ... [13:40] ir cdc_composite [13:40] *or [13:41] i dont think it is enabled in our kernel package [13:41] ok, i need to run [13:41] needs some config changes [13:41] will read that on my return [13:41] run lola run ... [13:41] :) [13:41] modprobe g_multi did find usb0 [13:41] oh [13:41] but then something about LUN0 [13:41] and bailed [13:41] well, there you go [13:41] yeah, likely needs modoptions [13:42] ogra_: http://pastebin.ubuntu.com/11815545/ [13:42] * Chipaca run s [13:42] (it can also be serial and fileservice) [13:42] yeah, figures [13:42] anyway! run! me! go! now! [13:42] :) [13:43] Chipaca: did you figure out that firstboot thing? [13:44] sergiusens, oh, regarding sabdfl's comment on the ML i guess we should seed fan again (still feels like bloating the image to me ... :) ) [13:45] ogra_: right, we should [13:45] i'll try to get to it before EOD [13:45] ogra_: I feel ashamed for conflating dhcp and dnsmasq though :-P [13:45] why [13:45] it acts as dhcp server in that setup [13:46] ogra_: because I could of nailed it down to a package [13:46] was just a glitch in terminology ... still matched the function ;) [13:47] elopio: are you here already? [13:47] is he working today ? [13:48] * ogra_ totally lost overview who is working and who isnt today :) [13:48] ogra_: hmm, Costa Rica isn't part of the US yet, right? :-P [13:48] wasnt there some religious day in SA today too ? [13:50] ogra_: not in the lower cone [13:51] ogra_: not according to http://www.timeanddate.com/holidays/costa-rica/ [13:51] ah, well :) [13:55] ogra_: I think most of the people not in the US are just taking swap days [13:56] yeah [13:57] lots of them though :) === dpm_ is now known as dpm [14:04] Chipaca: noupdategrub can land (did the 4 phased update test) [14:08] Chipaca: care to top approve https://code.launchpad.net/~sergiusens/snappy/noUpdateGrub/+merge/261381 ? === zyga_ is now known as zyga [14:37] good morning. [14:37] I'm having some internet problems today. [14:38] sergiusens: I'm here. [14:38] celebrating my independence by working from the office :) [14:38] sergiusens: how can I help you? [14:42] elopio: heh :-) [14:42] elopio: just wanted to have a quick talk on what the no update grub update implies wrt testing; also planned to invite fgimenez (and Chipaca) [14:42] makes for a pure spanish meeting [14:43] if not for ogra_ our standup today would of been in Spanish ;-) [14:43] lets not invite ogra_ to the standup either :) [14:47] well, it might need to happen after standup as I can't get a hold of Chipaca or fgimenez :-P [14:48] sergiusens, no, sorry :) i'm ok with having it now, looking forward to practice my spanish, i'm mostly forgetting it :) [15:03] sergiusens: back [15:25] Chipaca: pretty please approvez the branchez [15:25] @reviewlist [15:25] https://code.launchpad.net/~fgimenez/snappy/fix-failover-versions/+merge/263765 | No reviews (less than a day old) [15:25] https://code.launchpad.net/~mvo/snappy/snappy-fix-bbb-crash/+merge/263530 | No reviews (1 day old) [15:25] https://code.launchpad.net/~fgimenez/snappy/build-test/+merge/263327 | No reviews (2 days old) [15:26] https://code.launchpad.net/~fgimenez/snappy/filter-tests/+merge/263222 | Approve: 2 (3 days old) [15:26] https://code.launchpad.net/~mterry/snappy/selftest-reboot-notice/+merge/262265 | Approve: 1 (15 days old) [15:26] sergiusens: i suspect nothal is being a wuss [15:26] Chipaca: nothal got a flood warning I hope as this is missing: https://code.launchpad.net/~sergiusens/snappy/noUpdateGrub/+merge/261381 [15:27] sergiusens: yeah, i've got that one, but you used the plural [15:27] Chipaca: and all of these https://code.launchpad.net/~phablet-team/goget-ubuntu-touch/trunk/+activereviews which have the +1 but not the top +1 [15:27] sergiusens: so i was wondering what the other ones are :) [15:27] hah! [15:27] ok [15:28] @help reviewlist [15:28] Show the list of merge proposals with status = 'Needs review'. [15:28] usage: @reviewlist [project:] [user:] [15:29] @reviewlist user:sergiusens [15:29] https://code.launchpad.net/~fgimenez/snappy/fix-failover-versions/+merge/263765 | No reviews (less than a day old) [15:29]